The role of testosterone as an anabolic agent is widely misunderstood and the mechanisms of its anabolic effects on the muscle are largely unknown. Testosterone replacement increases fat-free mass in HIV-infected men with weight loss, but its efficacy in improving physical function has not been demonstrated. The objective of this competing continuation is to determine if testosterone replacement of HIV-infected men with weight loss will improve physical function, perceptions of physical function and body image. A second objective is to elucidate the mechanisms by which testosterone increases muscle mass; we will determine whether testosterone supplementation increases satellite cell number by promoting G:S phase transition, thus inducing satellite cells to enter the cell cycle or by inhibiting satellite cell apoptosis. We will also determine the effects of testosterone supplementation on muscle protein breakdown and protein synthetic efficiency. HIV-infected men, 18-60 yrs of age, with serum testosterone less than 350 ng/dL, and free of acute illness, will be randomly assigned to receive either placebo or 75- mg testosterone gel daily for 16 weeks. 75-mg testosterone gel, when applied daily, will raise serum testosterone in to the mid-normal range. Energy and protein intake, and exercise stimulus will be standardized. We will assess intramuscular molecular and cellular events and measure breakdown rates of mixed skeletal muscle protein by the combined use of the AV difference method and the three-pool model at baseline and after 2-weeks of treatment. The number of satellite cells and myonuclei will be assessed by electron microscopy, and satellite cell apoptosis by TUNEL method. Phosphorylated retinoblastoma protein and proliferating cell nuclear antigen (PCNA) will be used as markers of satellite cell cycle events. We will measure myostatin, IGF-1, and IGFBP-4 mRNA and protein concentrations by RT-PCR and Western blot analyses, and ubiquitin and proteasome mRNA by northern blots. The following outcomes will be measured at baseline and after 16 weeks: physical function by the stair climbing power, walking speed, and load carry test using threshold-independent methods; muscle performance by measurements of power, endurance, and 1 -repetition maximum strength in the leg press exercise; effort-independent muscle performance by force:EMG relationship; thigh muscle volume by MRI scan; perception of physical function, fatigue/energy, and body image by validated instruments. Total and free testosterone and DHT levels will be measured as markers of androgen bioavailability, and LH, FSH, and SHBG as markers of androgen action. For safety, we will follow hematocrit, AST and ALT, PSA, plasma lipids, apolipoproteins, and lipoprotein particles, digital rectal examinations, and sleep apnea questionnaire. A multi-disciplinary team of investigators, careful subject selection, access to a large patient pool, attention to potential confounding variables such as dietary intake, exercise stimulus, learning effect, and power and effect size, and state-of-the-art methods should maximize the chances of detecting treatment effects and elucidating the mechanism of androgen action. This study should help identify a therapeutic intervention that might improve physical function in HIV-infected men with weight loss, and enhance our understanding of the mechanisms by which testosterone stimulates muscle accretion.